Bioinformatics, 32(13), 2016, 2065–2066
doi: 10.1093/bioinformatics/btw096
Advance Access Publication Date: 22 February 2016
Applications Note
Systems biology
Reaction Decoder Tool (RDT): extracting
features from chemical reactions
Syed Asad Rahman1,2,*, Gilliean Torrance1, Lorenzo Baldacci1,3,
Sergio Martınez Cuesta1,4, Franz Fenninger1,5, Nimish Gopal1,
Saket Choudhary1, John W. May1,6, Gemma L. Holliday1,7,
Christoph Steinbeck1 and Janet M. Thornton1
1
European Molecular Biology Laboratory, European Bioinformatics Institute EMBL-EBI, Wellcome Trust Genome
Campus, Hinxton, Cambridge CB10 1SD, UK, 2Congenica Ltd, Wellcome Trust Genome Campus, Hinxton, Cambridge
CB10 1SA, UK, 3DISI, University of Bologna, V.Le Risorgimento 2, Bologna, Italy, 4Cancer Research UK Cambridge
Institute, University of Cambridge, Li Ka Shing Centre, Robinson Way, Cambridge CB2 0RE, UK, 5Michael Smith
Laboratories, the University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada, 6Innovation Centre
(Unit 23), Cambridge Science Park, NextMove Software Ltd, Cambridge CB4 0EY, UK and 7Bioengineering, UCSF
School of Pharmacy, San Francisco, CA 94158, USA
*To whom correspondence should be addressed.
Associate Editor: Alfonso Valencia
Received on November 9, 2015; revised on January 14, 2016; accepted on February 12, 2016
Abstract
Summary: Extracting chemical features like Atom–Atom Mapping (AAM), Bond Changes (BCs) and
Reaction Centres from biochemical reactions helps us understand the chemical composition of enzymatic reactions. Reaction Decoder is a robust command line tool, which performs this task with
high accuracy. It supports standard chemical input/output exchange formats i.e. RXN/SMILES,
computes AAM, highlights BCs and creates images of the mapped reaction. This aids in the analysis of metabolic pathways and the ability to perform comparative studies of chemical reactions
based on these features.
Availability and implementation: This software is implemented in Java, supported on Windows,
Linux and Mac OSX, and freely available at https://github.com/asad/ReactionDecoder
Contact: [email protected] or [email protected]
1 Introduction
Large-scale chemical reaction databases such as KEGG (Kanehisa
et al., 2013), BRENDA (Chang et al., 2015), Rhea (Alc
antara et al.,
2012) and MetaCyc (Latendresse et al., 2012) link reactions to enzymes and provide data-mining opportunities for novel pathways
(Hatzimanikatis et al., 2004; Rahman, 2007), and the discovery of
drugs, natural products and green chemistry. One of the primary
bottlenecks for automated analyses of these chemical reactions
comes from the realizations of the imperfect quality of data, such as
unmapped or unbalanced reactions. Accurate Atom–Atom Mapping
(AAM)—the one-to-one correspondence between the substrate and
product atoms (Gasteiger, 2003), will lead to correct prediction of
C The Author 2016. Published by Oxford University Press.
V
bond changes (BCs) (Rahman et al., 2014) and the ability to locate
the fate of interesting atoms or substructure across metabolic networks (Faulon and Bender, 2010) etc. Linking novel pathways or
optimizing pathways of biological/commercial relevance demands
better understanding of metabolic routes (Rahman, 2007) and pathway annotation (May et al., 2013).
We present Reaction Decoder Tool (RDT), a robust open source
software for mining reaction features, i.e. BCs, reaction centres and
to calculate similarity between reactions etc. The algorithm and
competing tools have been published in our previous article
(Rahman et al., 2014). Here we present the source code with relevant changes and below mentioned features.
2065
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
2066
2 Features
BCs in chemical reactions refers to the cleavage and formation of
chemical bonds, changes in bond order and stereo changes, which are
due to chemical processes such as chiral inversions or cis-trans isomerization(s). In the chemical reaction diagram and tables atoms are connected by a bond i.e. single ‘–‘, double ‘¼’or ring ‘%’ etc. (Fig. 1a),
for instance C–C means a single carbon–carbon bond that is cleaved
(‘j’) or formed in the reaction (‘k’). Bond order changes are represented by star ‘*’, e.g. C–C * C 5 C means a single carbon–carbon
bond turning into double carbon-carbon bond or vice versa. Stereo
changes are represented as atoms that change their absolute configuration, for instance C(R/S) means a carbon atom that changes from R
to S configuration. A reaction centre is the collection of atoms and
bonds that are changed during the reaction (Warr, 2014), also known
as the local atomic environment around the atoms involved in BCs.
The key features of this tool are:
i. Ability to perform AAM on chemical reactions catalyzed by enzymes (Fig. 1).
ii. In a balanced reaction, the total number of atoms on the left
side of the equation (Reactant), equals the total number of
atoms on the right (Product). In unbalanced reactions a best
‘guess’ is made.
iii. Generates images of the mapped reactions where matching substructures are highlighted.
iv. Generates reaction patterns and BCs for input reactions.
v. The input format and resulting mapped reaction with AAM information can be SMILES or RXN file (Gasteiger, 2003).
vi. The SMSD (Rahman et al., 2009) and CDK (Steinbeck et al.,
2006) are used to process chemical information.
3 Usage and applications
Tools like EC-Blast (Rahman et al., 2014), FunTree (Sillitoe and
Furnham, 2016), MACiE (Holliday et al., 2012) etc. use RDT in the
Fig. 1 (a) AAM performed by RDT and the resulting BCs (i.e. formed/cleaved:
C%C (Ring), Order Change: C%C*C ¼ C, C-C*C ¼ C) and (b) reaction centres
are highlighted in pink colour in the image, where as neighbouring atoms are
highlighted in green
S.Asad Rahman et al.
background to mine and extract chemical information from thousands of enzyme reactions. The success rate of mapping is >99%
when compared with manual AAM mappings (Rahman et al.,
2014). Originally developed to explore enzyme reactions, the tool is
also useful to explore any kind of organic chemical reaction
(Martınez Cuesta et al., 2014).
4 Conclusion
Reaction Decoder is a robust tool to compute AAM and extract
chemical features and calculate similarity between chemical reactions. This is coded in java and optimized to run as a computationally asynchronous process. It is distributed under GNU-GPL V3
license.
Acknowledgements
We would like to thank the CDK team, Sophie Therese Williams and anonymous users for their valuable feedback.
Funding
Authors would like to thank EMBL for funding this project. LB was partially
funded by Marie Curie fellowship.
Conflict of Interest: none declared.
References
Alc
antara,R. et al. (2012) Rhea—a manually curated resource of biochemical
reactions. Nucleic Acids Res., 40, D754–D760.
Chang,A. et al. (2015) BRENDA in 2015: exciting developments in its 25th
year of existence. Nucleic Acids Res., 43, D439–D446.
Faulon,J.L. and Bender,A. (2010) Handbook of Chemoinformatics
Algorithms. Chapman & Hall/CRC, Boca Raton, FL.
Gasteiger,J. (ed.) (2003) Applications, in Handbook of Chemoinformatics:
From Data to Knowledge in 4 Volumes. Wiley-VCH, Verlag GmbH,
Weinheim, Germany.
Hatzimanikatis,V. et al. (2004) Metabolic networks: enzyme function and metabolite structure. Curr. Opin. Struct. Biol., 14, 300–306.
Holliday,G.L. et al. (2012) MACiE: exploring the diversity of biochemical reactions. Nucleic Acids Res., 40, D783–D789.
Kanehisa,M. et al. (2013) Data, information, knowledge and principle: back
to metabolism in KEGG. Nucleic Acids Res., 42, D199–D205.
Latendresse,M. et al. (2012) Accurate atom-mapping computation for biochemical reactions. J. Chem. Inf. Model., 52, 2970–2982.
Martınez Cuesta,S. et al. (2014) The evolution of enzyme function in the isomerases. Curr. Opin. Struct. Biol., 26, 121–130.
May,J.W. et al. (2013) Metingear: a development environment for annotating
genome-scale metabolic models. Bioinformatics, 29, 2213–2215.
Rahman,S.A. (2007) Pathway hunter tool (pht)- a platform for metabolic network analysis and potential drug targeting. PhD Thesis, University of
Cologne, Cologne, Germany.
Rahman,S.A. et al. (2014) EC-BLAST: a tool to automatically search and compare enzyme reactions. Nat. Methods, 11, 171–174.
Rahman,S.A. et al. (2009) Small Molecule Subgraph Detector (SMSD) toolkit.
J. Cheminf, 1, 12–12.
Sillitoe,I. and Furnham,N. (2016) FunTree: advances in a resource for exploring and contextualising protein function evolution. Nucleic Acids Res., 44,
D317–D323.
Steinbeck,C. et al. (2006) Recent developments of the chemistry development
kit (CDK) - an open-source java library for chemo- and bioinformatics.
Curr. Pharm. Des., 12, 2111–2120.
Warr,W.A. (2014) A Short Review of Chemical Reaction Database Systems,
Computer-Aided Synthesis Design, Reaction Prediction and Synthetic
Feasibility. Mol. Inf., 33, 469–476.